Internet Disputes, Fairness
in Arbitration and
Transnationalism: A Reply to
Julia Hörnle
Thomas Schultz*I Introduction
Julia Hörnle is one of the world’s foremost scholars in the field of on-line dispute resolution (ODR), a small field that lies at the crossroads of Internet law and dispute settlement. With Cross-border Internet Dispute
Reso-lution, she has put together a major work in which she expounds her vision
of a fair system to resolve Internet disputes. To be sure, very few people could have been in a better position than Hörnle to engage in the daunt-ing task of defindaunt-ing the lineaments of justice in cyberspace: it requires a thorough understanding of technological feasibility, sociological realism, business acumen, legal dexterity and basic emotional common sense. Hörnle clearly displays all of these strengths in her treatment of the sub-ject-matter. But analytical precision and taking arguments to their logical conclusion occasionally yield to such higher values in her exposition. And at a number of junctures her arguments regrettably seem to be driven by conventional wisdoms prevalent in the ‘ODR Community’ – a community that in 15 years has not been able to produce a straightforwardly laud-able and working model of ODR, as the one or two current ODR systems
* Swiss National Science Foundation Ambizione Fellow, Graduate Institute of International and Devel-opment Studies; Senior Lecturer (Maître d’enseignement et de recherche), University of Geneva Law Faculty; Executive Director, Geneva Master in International Dispute Settlement (MIDS); Editorial Director, Journal of International Dispute Settlement. Research supported by the Swiss National Science Foundation. Email: Thomas.Schultz@graduateinstitute.ch.
that are living up to expectations rely on specificities that can barely be replicated elsewhere with any systematicity. (I also belong to this com-munity and share the blame.) At these junctures, one would have hoped someone of Hörnle’s calibre and acumen to be more critical, more in-novative, more challenging – more ‘hungry and foolish’ as the authors of the ‘Whole Earth Catalog’ put it in their hardcopy forerunner of Google. Nevertheless, the sweep and depth of Cross-border Internet Dispute Resolution definitely command the admiration of any reader interested in informa-tion technology and e-commerce. It is the most thoroughly researched, clearly presented and insightful book on online dispute resolution written so far - a position it is likely to keep for quite some time. It is a must-read for all practitioners and theorists of dispute resolution, but some of the fundamental ideas of the book call, I believe, for a debate which this art-icle offers to start.
II The importance of fairness
The book is divided into eight chapters. The titles of the second and last chapter (the first chapter being the introduction) are likely to make anyone interested in ODR forget about the kids to fetch at the crèche, out of sheer fascination for what they promise: ‘The concepts of fairness’ and ‘A model of dispute resolution for the Internet’. The quest for a fair (in the broad non-technical meaning of operational, satisfactory, popular) model of dispute resolution for the Internet has been the Grail Quest for thousands of Internet entrepreneurs, hundreds of arbitration and Internet lawyers, and a handful of legal philosophers for the last decade. Hörnle tackles the question head on. Indeed, her work is deeply anchored in the belief that ODR must be fair (or just, which she uses as synonyms). This central idea steers the entire book, and rightly so.
Such an enterprise is very commendable, and it is often attempted by the most distinguished scholars in other areas of international dispute settlement, as it requires both technical command of the dispute reso-lution mechanism in question and a good dose of reflective reasoning in legal and moral philosophy. Hörnle’s first chapter includes indeed an interesting, albeit brief, discussion of Rawls, Habermas and Fuller (though the more pregnant work of Fuller on the rule of law and predictability is unfortunately not addressed).
However, this discussion barely finds echo in the rest of the book; it is doubtful that these philosophers’ thoughts have had any real influence on the book’s arguments. Habermas’s contributions to deliberative dem-ocracy, for instance, have been applied in a variety of other works to the question of the law applicable in arbitration, and there would have been no place better than this book to consider the ‘fairness’ of ‘transnational law’ and how ODR should consequently be modelled.
III Accuracy
At certain, in my sense critical, junctures, references to the work of moral philosophers give way to arguments based on what appear to be moral gut feelings. For instance, when Hörnle argues that ‘a decision based on wrong facts is by definition unfair’ (p6), one is forced to recognize that she has not explored all the involuted ramifications of this idea. Indeed, if a decision were fair exclusively if the minor premise of the syllogism that produced it contained no wrong facts, then decision-makers would be bound to undertake an endless quest for perfect truth, in the sense idealized by Douglas Adams in the Hitchhiker’s Guide to the Galaxy: During a trial, a witness called Prak is given an overdose of a truth drug and then asked to ‘to tell the Truth, the Whole Truth and Nothing but the Truth’. In what is represented as an exceedingly long time after the trial, a jour-nalist covering the case meets one of the central characters of the book, who inquired about what happened next and was given the following an-swer: ‘“Oh, he told it all right,” said the man savagely, “for all I know he’s still telling it now”’. 2 Less metaphorically, William Park, in a recent study on truth-seeking in arbitration, writes that ‘truth-seeking balances against sensitivity to speed and economy in arbitration. Accuracy in arbitration means something other than absolute truth as it might exist in the eyes of an omniscient God . . . [A]rbitrators aim to get as near as reasonably pos-sible to a correct picture of [reality].’3
Speed and economy are particularly sensitive in online arbitration. In fact, the very raison d’être of online arbitration is the parties’ pursuit of a radically faster and cheaper form of dispute resolution. The disad-vantages inherent in the replacement of face-to-face communication by online communication are only acceptable as a trade-off for significantly increased speed and drastically reduced costs. Given as much, truth must yield to efficacy: as the adage aptly cited by Park says, veritas filia temporis (truth is the daughter of time).4 In the field of arbitration, truth is the daughter of time and costs. When arbitrating a dispute worth for instance EUR 2000 (which is a rather large dispute in the world of Internet con-sumer disputes), the constraints on time and costs are not mere irksome idealized requirements, they are strict cut-offs on the ladder of feasibility. If a EUR 2000 dispute generates more than a few hundred Euros in costs when submitted to arbitration, it will simply not be submitted to arbitra-tion. It seems beyond the realm of realistic possibility that truth be estab-lished with such accuracy and certainty as to guarantee that no wrong
2 D. Adams, A Hitchiker’s Guide to the Galaxy: Life, the Universe, and Everything (Pan Books, London, 1979),
chapter 34.
3 William W. Park, ‘Arbitrators and Accuracy’ (2010) Journal of International Dispute Settlement 25, 26. 4 Ibid, 27.
fact forms part of the basis for the decision, while generating costs of no more than a few hundred Euros in costs. Approximate decision-making, in the sense of a rough correspondence between the facts in the deci-sion and the facts in reality, is the best we can get in online arbitration of small cross-border disputes. For such disputes, arbitration is no longer the truth-seeking process that it is for commercial, investment or interstate disputes,5 but a process to avoid crass disrespect of the contract or basic legal obligations in a consumer transaction.
IV Due process
This somewhat idealized notion of fairness as near-perfect accuracy in decision-making appears to radiate throughout the entire book. In the heart of the work, located in Chapter 7, Hörnle tells us that ‘stricter due process standards should be applied to online arbitration of Internet dis-putes . . . than have hitherto been applied to offline commercial arbitra-tion between businesses’ (p171). To sustain this claim, Hörnle appeals to the parties’ (and especially the weaker party’s) lack of alternatives, as access to courts is forbiddingly expensive given the value of the dispute (which the author aptly addresses in Chapter 3), and to the power im-balance that typically characterizes small Internet disputes. In theory, it is likely that most proponents of equality before the law would embrace this idea. However, we can probably dispense with the philosophical analysis of the need for additional due process requirements imposed by fairness in the resolution of Internet disputes. The same reasons apply here that led, in the preceding paragraphs, to the acceptability of a relatively low level of correspondence between the facts in the decision and the facts in reality.
At this juncture, the Court of Arbitration for Sport (CAS) comes to mind. Athletes who want to participate, for example, in the Olympic Games, must enter into an arbitration clause in order to do so, and there frequently is a significant power imbalance between the parties, that is the athlete and the sports federation. The situation is similar to the one envis-aged by Hörnle. Arbitral tribunals of the ad hoc Division of CAS, which deal with disputes arising during the Olympic Games or up to ten days before the Games, are required to give decisions within, in principle, 24 hours of the lodging of the application.6 The parties consequently have only limited possibilities to be heard, which translates as lower due process standards.7 Under Hörnle’s analysis, such arbitrations would seem quite
5 Ibid.
6 Article 18 of the Arbitration Rules for the Olympic Games.
7 Thomas Schultz, ‘Human Rights: A Speed Bump for Arbitral Procedures? An Exploration of
unfair. In reality, it is the only practicable option: if an arbitral tribunal decided to either exclude or readmit an athlete to the Games by an award rendered after the end of the Games, the dispute resolution mechanism would barely get any praise for its justice. Justice would remain elusive regardless how much fairness one would hope to infer from extended hearings and the hence increased level of due process. As Voltaire said, the best is the enemy of the good – this applies to sports arbitration during the Olympics and to online arbitration dealing with small cases.
Hörnle recognizes this difficulty, which she labels the ‘conflict between due process and effectiveness’ (p249), but the consequences she attaches to it appear too modest. First of all, she sets a de minimis for disputes amen-able to online arbitration: according to the author, disputes of GBP 100 and above may rightfully be submitted to online arbitration (p250f). Hörnle gets to this threshold value after a logical lapse: she first poses the question ‘whether disputes below a certain value can be solved at all by online arbitration that satisfies the [standards of] procedural due process [that the author considers appropriate]’ (p250). Answering this question in the negative (there is a threshold value below which it is financially un-viable to submit a dispute to arbitration), she proceeds to determine this value by reference to ‘the amount any household can afford without sub-stantially damaging its standard of living’ (p251). This constitutes a con-fusion between a normative statement (online arbitration should not be bothered with certain trifles, which we could prescriptively define as issues below a value of GBP 100) and a descriptive statement (it is not financially viable for online arbitration to deal with certain disputes). This logical lapse cuts to the heart of the problem: One may readily agree with Hörn-le’s conclusion that disputes below GBP 100 should not go to online arbi-tration for reasons of proportionality (we should not bother arbitrating a dispute worth GBP 99, for instance). But the real problem is the threshold value below which online arbitration is no longer financially viable (what, in Hörnle’s words, ‘can be solved at all by arbitration’). Can we arbitrate a GBP 100 dispute? Can we arbitrate 100’000 disputes worth GBP 100, or GBP 500, or GBP 1’000? Her eschewal of this problem, which is posed but not answered, may precisely be what allows her to proceed to her next and main argument in the ‘conflict between due process and effectiveness’.
Hörnle considers that the appropriate model for online arbitration is ‘a model of adjudication that (in terms of due process) is positioned be-tween litigation and commercial arbitration’ (p250). In other words, for Hörnle the level of due process in online arbitration should be higher, for disputes worth as little as GBP 100, than in commercial arbitration, where disputes are typically between 1000 and 100 000 times higher.
More precisely, for Hörnle this means that an arbitral tribunal which has to decide on a case worth for instance GBP 100 has to apply not only the lex arbitri (p95ff) and the relevant institutional procedural rules (if any), but also specific standards for online arbitration (for example EC
Recommendation 98/257/EC, the Guidelines of the ABA Task Force on E-Commerce and ADR, the UK Office of Fair Trading’s Consumer Codes Approval Scheme, etc) (p93ff) and human rights standards such as those enshrined in the European Convention on Human Rights (p98ff). These different layers of procedural requirements should be interpreted in such way that they impose ‘higher standards of due process’ (p168) and accordingly greater procedural constraints and duties on the arbitrator dealing with a GBP 100 Internet dispute than on the arbitrator dealing with a construction dispute worth a few million Pounds. Following a wild guess, I would rather have been tempted to fit ODR under the 80-20 rule – the management principle roughly linked to Pareto which provides that 80% of the results come from 20% of the efforts. However unquantifiable due process is, it may have been interesting to examine the following: if an arbitrator made 20% of the procedural efforts in online arbitration she would normally make in an international arbitration, would she not achieve a level of due process roughly comparable to 80% of the due pro-cess usually found in ‘ordinary’ international arbitration? Imperfect as it is, would such a solution not have been reasonably acceptable on the plane of justice and infinitely more workable in practice than to ask an arbitrator in a small case to be even more cautious about due process than in a large arbitration?
At this juncture, it may appear useful to ponder the reasons an arbi-trator may have to try to meet the elevated due process standards that Hörnle suggests, in the context of more general reflections of what is likely to happen in practice. A conceptual instrument that may be useful in this regard is the distinction between prudential reasons-for-action and moral reasons-for-action. Reasons-for-action are the ‘factors that [motivate or] would motivate [people] if they were to understand the serviceability of those factors for the furtherance of their general objectives.’8 Pruden-tial reasons-for-action are typically opposed to moral reasons-for-action, in that the former rely on objective interests, on objective gains and losses associated with different courses of action, whereas the latter rely on the morality of different courses of action.9
It is possible, even likely, that an arbitrator would find moral reasons to apply the heightened due process standards, advocated by Hörnle, in an online arbitration of an Internet dispute because of the possible lack of true consent to arbitrate and the power imbalances between the parties. Prudential reasons-for-action, however, create the opposite incentives, supporting a low due process standard. First, online arbitration is used because access to court is forbiddingly expensive for small cross-border
8 Matthew H. Kramer, ‘On the Moral Status of the Rule of Law’ (2004) 64 Cambridge Law Journal 65, 67. 9 Joseph Raz, Practical Reason and Norms (OUP, Oxford, 1999), 155–156; Matthew H. Kramer, In Defense
cases. This lack of access to the court also implies that an ‘online award’ cannot realistically be challenged in court. An arbitrator applying lower standards of due process than those required by the lex arbitri is very un-likely to see her award set aside in court. Second, it seems very difficult to create the right financial incentives, with a case worth GBP 100 or even 20 times that value, to lead an arbitrator to examine, for instance, what exactly the European Convention on Human Rights may entail for arbi-tral proceedings, as Hörnle would seem to require.
For parties to small cross-border disputes, there seem to be only two choices: no justice or rough justice, the sort of rough justice found in simplified arbitration procedures and in small claims courts. The kind of elaborate and admirable form of fairness (let us abstain from using more tempting adjectives such as ‘Byzantine’) we find in the meanders of human rights standards in general and the case-law of the European Court of Human Rights in particular do not, from a systems design point of view, appear to be the ingredient that would make ODR the success that many analysts had hoped it would become. To resist, as Hörnle does, ‘[v]ery tight and inflexible word limits [and] deadlines for filing submis-sions’ and ‘strict limit[s] on the number of submissions by the parties’ (p242) may be the consequence of very sound moral considerations, but from a prudential point of view, it does not seem to be a commendable solution. The same considerations apply to Hörnle’s call for a double level of arbitration (which she curiously labels ‘Judicial review/appeal’ – p244), with appeals to a higher level of arbitration on ‘important points of law and for procedural challenges’ (p244). Hörnle seems to encourage us to imagine an arbitration of a case worth for example EUR 1000, in which the applicability and precise implications of the European Convention on Human Rights on arbitration are debated, with no tight and inflexible word limits and no strict limit on the number of submissions. This case and the debated questions would then further be appealed to a higher instance of arbitration. Such a hypothetical testifies to an idealism that I am not ready to embrace.
V Applicable law
Finally, after fact-finding and due process, the same need to ‘calibrate the expenditure of time and money’, as Park puts it,10 affects the question of the law applicable in online arbitration. Hörnle’s basic tenet is that ‘Arbitration fulfils the same function as court proceedings, i.e. adjudicating a dispute according to laws and legal rules’ (p69) and ‘there should be no discrimination between litigation and arbitration, since both forms
of adjudication fulfil a public function’ (p70). Accordingly, she consid-ers that the determination of the applicable law in arbitration should in essence follow the model of courts. She contends in particular that ‘party autonomy [in the international arbitration of Internet disputes] should be restricted to the extent that mandatory rules [in the international sense] apply . . . if the conditions for their application are fulfilled’ (p70). This, she considers, is a question of justice: the weaker party should not be deprived of the lois de police designed to make up for its weakness (p65f). Fair enough. But would this not be a Pyrrhic victory for consumers? Hörnle’s position would lead us to this: a EUR 500 cross-border Internet dispute goes to arbitration, because courts are no alternative due to the costs that they generate; the arbitrator is asked - regardless of the par-ties’ consent to see their dispute resolved in application of a given set of rules - to examine whether she should apply the weaker party’s national consumer protection law; in order to make this determination, the arbi-trator will have to examine, for instance, if the weaker party can be con-sidered a consumer or not (should the definition of ‘consumer’ be wider for Internet disputes, as Hörnle contends at p32-37?), if the business has directed its activities to the country of the consumer (when, exactly, is a website directed to a specific country?), and what the national consumer protection law exactly provides (what, exactly, are the contents of, say, Danish consumer law?). Regardless how the arbitrator’s fees are set and funded, this does not seem appropriate for a dispute worth the equivalent of one or two hours’ work by an arbitrator. The problem, I contend, lies in the epistemological obstacle created by the political ideology underlying Hörnle’s tenet that dispute resolution is fundamentally a public function (p.217), and in the tension that this ideology creates between the rule of law and efficiency.
The political ideology that can be ascribed to Hörnle’s thinking is that of classical legal positivism (roughly speaking, legal positivism prior to the Hart-Fuller debate). The gist of this ideology was formulated by Léon Duguit and Hans Kelsen in the following words: ‘we believe to have ser-ious reasons to be convinced that the only means to satisfy our aspiration for justice and equity is the resigned confidence that there is no other justice than the justice to be found in the positive law of states’.11 Anyone embracing this ideology, consciously or unconsciously, is bound to come to a specific set of answers when asking the fundamental question in the issue of the applicable law: Where can we look, when in search of justice? The answer, with such an ideology, can only be a national law. Assum-ing, arguendo, that the only choice is between the law of the consumer
11 Léon Duguit and Hans Kelsen, Foreword (1926-1927) 1 Revue internationale de la théorie du droit 3,
(or more generally the weaker party) and the law of the business, then it seems intuitively fair to conclude in favour of the former.
Let us now ponder the question of the applicable law without this epis-temological obstacle, taking into consideration the possibility that there could be justice outside of state law. The concept of meta-justice may be called upon at this juncture. Recently introduced into the arsenal of pri-vate international lawyers by Alex Mills, meta-justice refers to the justice of the principles governing the global ordering of legal systems (as systems of justice, hence the word ‘meta’ in meta-justice) that private international law embodies.12 Meta-justice determines the justice (or fairness, in the ter-minology used by Hörnle) of the shapes given to the regulatory scope of different legal systems. It asks not whether the substantive result provided by a legal system for any given case is more just than that provided by any other legal system, but proceeds on a higher level of abstraction. It asks whether it is just to grant a certain legal system regulatory authority over a given situation. One of the main criteria of such justice, thus ascribable to the determination of the applicable law, may perhaps be predictability. Indeed, predictability is the most fundamental pursuit of the rule of law.
Predictability is the fundamental goal of the rule of law, both as a moral-political ideal and as the manifestation of a jurisprudential concept (the state of affairs that obtains when the necessary and sufficient conditions of legality are fulfilled). Brutally simplified, the rule of law, when it is under-stood as a moral-political ideal, consists of a set of precepts of political mor-ality that essentially seek to guarantee a high level of predictability; when the rule of law is understood as the presence of a functional legal system, it sets forth conditions for the existence of a regime of law that are in essence various facets of predictability.13 Put differently still, it is more fundamen-tally important, for a regime that aspires to the sort of justice that we asso-ciate with law, to be globally predictable than to provide substantively just but unpredictable results for individual cases. Predictability is one of the main values that emerge when we shift the focus from fairness of the in-stant case to systemic fairness, and systemic fairness lies closer to the heart of the concept of law than the substantive fairness of individual cases.14
Now, predictability can be provided by law beyond the state (a non-state normative order), sometimes more so than by national law. Let us
12 Alex Mills, The Confluence of Private and Public International Law (CUP, Cambridge, 2009) 18: ‘The
distribution of regulatory authority is itself a question of justice – perhaps meta-justice - operating within the framework of justice pluralism. The answer to a problem of private international law involves the determination of which idea of justice (which state law) would be most just to apply. Private international law, properly understood, is about determining the most “just” distribution of regulatory authority.’ Note Mills’s focus on state law, a limitation that is not recognized by the present author.
13 Matthew H. Kramer, Objectivity and the Rule of Law (CUP, Cambridge, 2007).
14 See further Thomas Schultz, ‘The Three Pursuits of Dispute Settlement’ (2010) 1 Czech & Central
imagine an online trading platform, very much transnationalized in the sense that the location of buyers and sellers is almost meaningless: accord-ing to which rules would the users of this platform expect (that is, pre-dict) that their disputes will be resolved? This theoretical question was tested almost ten years ago in a real setting: eBay’s online marketplace. The result of the experiment was that eBay users expected their disputes to be resolved according to ‘eBay Law’ (the various norms, contractual and social, that were created by eBay and the community of eBay users) and not according to any national, consumer law.15 Indeed, it is infinitely more predictable for eBay users that their disputes are resolved according to a single set of transnational rules, rather than to myriad consumer laws, whose workings typically require a level of interpretative proficiency that is beyond the reach of the normal consumer. The same reasoning applies for instance to domain names: the application of the rules of the Uniform Domain Name Dispute Resolution Policy (UDRP) is incomparably more predictable than the application of variegated national trademark laws.16 Transnationalism – law beyond the state – may be the key to predictability, and thus to the sort of justice, or fairness, that is central to the rule of law. In this sense, the UDRP and ‘eBay law’ are more just than national laws, because the former more easily allow the rules’ addressees to plan their affairs and predict their consequences.17
It may further be pointed out that the application of a single set of transnational rules is easier, less costly and faster – in other words more efficient - than applying national consumer law with all the intricacies of legal reasoning that this involves. The parties to small cases may well prefer to get just some predictable remedy, rather than one that requires fascinating intellectual challenges, which we academics so often tend to idealize and favour.
The focus of any ODR system aspiring to the sort of justice that we as-sociate with law should thus not be on correctly applying the (national) law on point, but on developing and applying a set of predictable trans-national rules. And its focus should be on systemic fairness rather than on the fairness of the resolution of the instant case. Perhaps the fairness of the resolution of an instant case would be greater following Hörnle’s approach, but transnationalism would increase the fairness of the system, by making it more predictable, less costly, thus more workable and pro-viding greater access. To make that shift of focus, we need to embrace the
15 Ethan Katsh, Janet Rifkin and Alan Gaitenby, ‘“eBay Law”’ (2000) 15 Ohio State Journal of Dispute
Resolution 705.
16 See further Thomas Schultz, ‘Private Legal Systems: What Cyberspace Might Teach Legal Theorists’
(2007) 10 Yale Journal of Law & Technology 151.
17 See Brian Z. Tamanaha, On the Rule of Law (CUP, Cambridge, 2004) 96, who speaks of ‘enhanc[ing]
adage that it is more important for a rule to be settled than it be settled right. It is such an approach that would reconcile the rule of law and effi-ciency.
6 Conclusion
The level of disagreement of the present author with Hörnle’s wonderful treatment of the subject-matter does not carry any criticism on the value of Hörnle’s book. The questions addressed here are genuine issues, not matters of confusion. Nevertheless, all in all, Julia Hörnle’s book is a clarion call for morality in online dispute resolution in general and in on-line arbitration in particular. Commendable as it is, this call should not be answered: to echo Confucius, we should not use the cannons of human rights standards and mandatory rules in the international sense to kill the flies of disputes worth a few hundred Euros.
